DE 101 44 076 A1 discloses a device and a method for the early detection and prediction of damage to the assemblies of a machine system. For this purpose, the solid-borne sound of the machine system is detected by a sensor, output as an acceleration signal and analyzed in a digital signal processor. In order to avoid adverse influences of vibrations in the surroundings and of solid-borne sound waves which are not associated with the state of the machine system, the acceleration signal is firstly transformed into the frequency range by means of a fast Fourier transformation. The data which are obtained as a result are then transformed into the time domain again by means of a cepstrum analysis, with the result that resonance data are obtained from individual pulses in the time domain. This cepstrum is then compared with a comparison cepstrum which is correspondingly available to load signals and rotational speed signals for the present operating state in a new machine system in a storage device. When limiting values are exceeded, the diagnosis signal, in particular information about the assembly which is diagnosed as being damaged and its predicted remaining service life, are displayed to the user and an emergency operation is initiated.